What Does It Take to Be a Good Printer?

By Ctein

One of the interesting (because it was unexpected) results of the educational print offer of a few weeks back was that people thought I invoked extraordinary measures to come up with a good print of that photograph. Not even close. I worked with a level of care and attention to detail and meticulousness that most people wouldn't muster, but that's a different thing. In that respect, I am unusual, and that's why I get paid for what I do.

But the measures themselves had nothing extraordinary about them. They're just the kind of things that any more-than-routine printer should be doing.

Good printing is not accomplished merely by getting curves and color balance and saturation correct. That's basic, like figuring out the right exposure and paper grade (or filter pack) in the darkroom.

One step above that is local tone correction—dodging and burning in. It's always amazed me (not in a good way) how many printers don't do that, in the darkroom or out. The majority of photographs I've printed over the past 40 years, and that's many thousands, have been improved by some amount of local correction. Frequently, it's very slight and very subtle. But it's not the norm for a print to be better with none at all.

If you don't have that under your control, you're not even a competent printer. If you are, you're competent. You can be trusted to turn out a good journeyman print.

Note: In each of the following pairs of photographs, the "before" version is what I get from my default settings for Adobe Camera Raw prior to making any adjustments to the photograph. The "after" version is what I use to make my finished prints; it isn't optimized for viewing at small resolution on a screen, but you'll get the idea. To put it another way, the "before version" is what the camera saw, and the "after" version is what I saw when I made the photograph—and what I wanted in a print.

Before (above) and after (below): The most important change to this photograph was a lot of dodging and burning in. Human beings are really good at ignoring large-scale variations in brightness that cameras see all too clearly.

What is it take to move beyond that level? Just as in the darkroom, it requires some real understanding of the limitations and quirks of the media (both film/sensor and print) you're working with. That leads me to three tools that are part of my standard printmaking toolkit. Some variation on these ought to be part of yours.

1. Local contrast enhancement. Simply, you can do it with a mild dose of unsharp masking or you can do it in a much more sophisticated way with plug-ins like Photo Wiz's ContrastMaster. What you can't get away with is not doing it. Digital prints tend to come out flat, with poor separation of fine and subtle tonal differences. I don't know why, but I've observed it to be broadly true. The delicate nuances that are there in the original file, whether it comes from a digital camera or a film scan, don't get rendered well in the print. You need to make that tonal separation stronger in the file or the print will look duller than it should.

Before (above) and after (below): The near-setting sun produced a grazing light that brought out wonderful textures in the building and provided great contrast in tone and color between the building in the sky. I made adjustments that emphasized sharpness and fine tonal detail.

2. Deconvolution sharpening. Simply, there's Smart Sharpen in Photoshop, with a radius of a few tenths of the pixel and the percentage set to taste. Complexly, any of several plug-ins, like Topaz Lab's InFocus. If you're trying to produce a digital photograph with the best and most nicely-rendered fine detail, and you're using a Bayer array camera, this is pretty much a must. (Can't speak to non-Bayer array cameras; never printed from them.) There is a genuine improvement in sharpness and fine detail buried in those Raw files that deconvolution sharpening can dig out.

3. Noise reduction. It's the counterbalance to the previous two tools. Both local enhancement and sharpening increase noise at the same time they improve fine detail, gradation, and texture. Noise reduction, on the other hand, tends to suppress precisely those image attributes. Many digital files (and film scans) are improved by just a whisper of noise reduction, no matter how great the camera is. A lot more digital files need it when local enhancement or sharpening are coming into play.

Before (above) and after (below): To turn this photograph from what the camera saw into what I saw, I needed a certain amount of dodging and burning in but far more important was bringing up the large-scale tonal variations so that they would separate and print well.

Combining all three of these in different measures gives you broad control over the characteristics image. You can get reduced noise with no visible sacrifice of fine detail and texture. Or, you can get substantially improved sharpness and detail with no significant increase in noise. Or, you can split the difference and get a photograph that has lower noise and better sharpness at the same time.

Before (above) and after (below): This photograph got "the works"; everything in my basic toolkit was important to making this photograph work as a print: Proper dodging and burning in, local contrast enhancement to bring out the gorgeous ripples in the water I had seen, proper sharpening to bring out the details in the seals and make them stand out against the water, and noise reduction to get rid of the increased noise from the previous operations.

None of this is heroic; none of it is exceptional. It's the norm for good digital printing.

No mincing of words: this is what it takes to be better than run-of-the-mill.

Ctein

Ctein's weekly column, which usually appears on Wednesdays, is a day late this week through no fault of his own.

[Note that the illustrations are only approximations. Real corrections would have to take size into account, as well as the big jump between screen and paper, and the TypePad blogging software has an effect on how illustrations look that is impossible to compensate for precisely. Finally, your monitor and its calibration might change what you're seeing relative to what Ctein—or anyone else—is seeing. —MJ]

Send this post to a friend

Please help support TOP by patronizing our sponsors B&H Photo and Amazon

Note: Links in this post may be to our affiliates; sales through affiliate links may benefit this site. More...
Original contents copyright 2012 by Michael C. Johnston and/or the bylined author. All Rights Reserved.

Featured Comment by Richard Tugwell: "I bought one of Charlie Cramer's prints from TOP a while back—quite beautiful. I see he has an interesting and in-depth article on tonal adjustments using Lightroom 4 over at The Luminous-Landscape."

Featured Comment by Ned: "In my opinion this is an area where books, videos and other training techniques just don't cut it. I don't even know what a good print looks like, which is why I purchased Ctien's Micro 4/3 print. I don't like to print because I know it will be a piece of crap."

Mike replies: Sounds like you might be a candidte to become a client of a custom printer. Don't forget that there are numerous photographers, now and in history—including some of the greats—who let other people print their pictures for them. It's how Ctein makes his living. Perhaps his best-known client was the late rock-and-roll photographer Jim Marshall, but he prints for many other clients as well, some of whom are quite famous. Years ago I made my living as a custom printer too—and many of my clients at the time never made their own prints.

Remember this picture?

It was taken by Peter Turnley at Voja Mitrovic's retirement party, and shows Henri Cartier-Bresson and Josef Koudelka good-naturedly begging Voja not to retire.

Featured Comment by Ed: "Mornin', Ctein, great read as ever. To me printing is about effort and the ability to reject (no matter the cost). In a German documentary I saw Gursky at work at Grieger (one of the best digital print shops on the face of this planet). He was making a test print...2 meters 60 cm by .5 meter. Just to check local contrast, after he rejected a full print of 'Hamm Bergwerg Ost.' If your print does not provide what you want, tear it up—it is a painful, wasteful, yet very important part of the process of learning to print. And looking at examples like the ones you provided, well, that may make people wonder about their skills (and their preferences), but, hey, you don't get 40 years of experience overnight."

Comments

[Hendrik]

It would be very kind and much appreciated, if someone could point out, whether the sort of adjustments/ processing Ctein talks about, is possible in Lightroom at all. And maybe add the corresponding names of tools there. Thanks.

Hendrik

[Edward Richards]

I use the other techniques, but deconvolution sharpening is new to me. (But I mostly work from LF negative scans, and I usually have plenty of detail.) Would you do deconvolution sharpening after local contrast enhancement, then final print sharpening last? Would you still do the initial capture sharpening, or do these on the unsharpened raw files?

[John Krumm]

Ctein, what do you think about us less than ambitious folks who do most of their local adjustments in Lightroom with the brush or grad tools? I know the masking is not as precise as what you can do in Photoshop (I would never attempt a mask like you made with your bridge) but I've been pretty happy with the results overall. Unfortunately I find it hard to print directly from Lightroom as the prints come out darker than I want, so I use Qimage for that.

[Ben Rosengart]

Is there any equivalent of Smart Sharpen in Lightroom?

[Daniel S.]

I see many references to "what [you] saw" when deciding what adjustments to make, which leads me to my question: how do you think your ability to make a good print from a given photograph changes with time? and how long until you see a definite effect?

[Craig C.]

Ctein,

As an owner of the 1st edition of your Digital Restoration book and owner of prints you've offered through TOP sales, I already know directly of your skills and knowledge as a master printer.

My question is whether you have plans for a book explaining what you look for and how you subsequently post process your images preparing them for print?

If not a book, have you considered holding a printing workshop?

As one of the unwashed masses, I'm not expecting to become anywhere near your skills as a printer, but learning some fundamentals from a master would start me on a path towards a much better state of satisfaction with the end results of my prints.

Best Regards,

[Eugene]

I`d argue that moderation with local contrast enhancement is key, to my taste both the building and the submarine photos brought here
feel somewhat 'overdone' and suffer from artifacts typical for local-contrast enhancement, that is bright hallos spilling on dark surfaces
and "dirt smudges" on white ones. Unless these go away in print I`d consider dialing it down.

[Bahi]

A note for those who've been asking about sharpening in Lightroom.

The Detail slider in Lightroom (and ACR) next to the Sharpening slider offers deconvolution sharpening (with the slider set at 100) and a less aggressive algorithm (one that's less likely to show halos and other sharpening artefacts) as the slider value approaches 0. In between, the sharpening used is a mix of the two methods, depending on slider position.

The local sharpening brush in LR uses these values but to combine different local sharpening methods based on different Detail settings or different radius settings, create virtual copies, alter the slider settings, apply sharpening as required and ask LR to open these all these virtual copies as layers of a single Photoshop document. (Select the images, right-click, Edit in-> Open as Layers in Photoshop.) You can then use layer masks and blends to combine the layers. When you save the multi-layer TIFF, it'll appear in Lightroom as a new file.

More info here.

Side note: it's worth reading everything that Eric Chan of Adobe (aka MadManChan) has to say about Lightroom and ACR.

[Marcelo Guarini]

I use Photoshop since I believe I can have more control than with Lightroom.

Thanks very much Ctein for sharing your knowledge.

Regards

[Steven Willard]

I've had this notion that most of us would be better off if we chased the kind of knowledge exhibited here in Ctein's article with the same enthusiasm we chase after the latest new gear. If we spent just a fraction of our budget on good seminars and classes we would probabley see more improvement in our output at less cost. The side benifit is that the knowledge is likely to last longer, and in the end be more rewarding. Just an idea.

[Will Frostmill]

Ctein,
Thank you for showing us how it's done. That is very helpful. In fact, like Ned, I bought one of your prints as a reference, to see what really good printing looks like.

It is interesting, all of your work feels like it is yours - and now I understand: it is because you are a good printer that you can make your photographs look not just like photographs, but like your vision. They reflect how you see.

In part, I can tell, because they aren't to my taste in the precisely the same way, but I cannot verbalize what that way is. What I mean is, they are excellent photographs, but they don't suit me, but they don't suit me the same way; there's no randomness to it. I hope you do not take offense?

I am envious, because often I can make beautiful pictures, but reflecting what I saw, or what I wish to see is hit or miss. Hopefully, this advice will fix that.

Will

[Stan B.]

Fortunately, Voja had pretty much his entire lfetime to hone his skills- today, many of us have to constantly relearn the fine points as editing software and the printers themselves, constantly change and evolve.

[Kenneth Tanaka]

@ Ned: "I don't even know what a good print looks like, ..."

You've raised a substantial issue, Ned, one that I'm sure many (maybe most) TOP readers actually share.

Let me make the matter easy for you: A "good" print is one that you enjoy looking at.

That's a sufficient definition for most people's purposes.

More refined definitions begin, and end, with the nature and intentions of the image. That, by the way, is contrary to the Internet photo crowd's techie notion that "dynamic range" and sharpness rule the scales. They don't.

Not to turn this comment into a treatise but let me show one vivid, but by no means extraordinary, example of my point.

One of the hallmarks of many works by the late, great photographer Roy DeCarava was his ability to crush a negative to within an inch of its legibility and then revive just enough to create truly magical images, such as the one below.

Roy DeCarava, Dancers, uptown “club dance,” New York, 1956/1981

By Internet standards this print sucks. But by art standards its a knock-you-on-your-ass fabulous print (and nearly impossible to buy at any price).

Living with a print for a while is, in my opinion, the best strategy for deciding what's "best" for the image. Simple images, such as Ctein's bridge, don't really require such contemplative gestation. Print, dry, give them a day or two to decide if they're good, repeat if needed.

But prints featuring more complex intentions should be allowed more consideration. My own solution is to print such images, usually at close to their primary size, and just tack them up to a magnetic wall in my office where I can see them every day under all kinds of light.

Mag Wall

Some of these prints have been up there for months.

So try that method, assuming you can reach a point where you can produce at least one print that you might consider acceptable.

But always, always judge "print quality" against the image and your intentions. Outstanding technical prints of crappy imagery are worthless. Ditto the inverse.

[Charles]

The best technique for local contrast enhancement that I've found uses a high-pass filter.
1. Extract the green channel of the image
2. Apply a high-pass filter at about radius 3
3. Filter the original image through the high-pass overlay at cautious strength, usually around 25 percent
This recipe is taken from operations available in Picture Window Pro. I assume PS has an equivalent procedure.

And for deconvolution sharpening, the free-or-make-a-donation program Raw Therapee 4 has it. You can apply have RT apply it to a TIF file if you are committed to a different raw converter.

[Richard Man]

If you print B&W, I cannot recommend George DeWolfe's book highly enough. It covers the same topics as Ctein here, but for B&W.

It makes my images pop so much more.

[Richard Man]

To answer the Lightroom questions - most of the adjustments mentioned in George DeWolfe's book are in fact LR adjustments, with a couple "last steps" that is done in Photoshop if you want that last pop.

[ctein]

Dear folks,

I have no experience at all with Lightroom, so I will leave others to answer specific questions about that program. Sorry I can't be of more help!

General comments, though: exotic masking like I did on the bridge photograph is rarely necessary. Very few of my printing efforts involve carefully-constructed masks. (I believe there are third-party plug-ins that will help Lightroom with masking, but I'm not positive.) You can be a very good printer of most of your photographs without getting into exotic masking.

What you do need, though, is the ability to exercise local control with some precision and in some detail. I know that early versions of Lightroom (and Aperture) did not much allow for that. You can't be a good printer without those local controls. So if that's your objective and you're using some ancient flavor of Lightroom, upgrading would be in order. Otherwise you're trying to use a screwdriver to drive nails.

~~~~~~

Dear Daniel,

Not entirely sure I understood your questions, but I'll give it a try. I get better and better every year. It's still true after 40+ years of high-end printing. It's my biggest objection to issuing photographs in limited editions. I was a much better printer in 1985 than I was in 1970, I was a much better printer in 2000 than I was in 1985, and I'm a much better printer today than I was at the turn of the millennium.

How long before you see a definite improvement in your printing? Well, at a minimum, it will probably take you a couple of weeks of fiddling around before you start to internalize the tools well enough to use them really effectively. Then figure about six months of serious work before you'll feel like you've really got it down pat. Which you won't, that's an illusion. See the previous paragraph.

pax \ Ctein
[ Please excuse any word-salad. MacSpeech in training! ]
======================================
-- Ctein's Online Gallery http://ctein.com 
-- Digital Restorations http://photo-repair.com 
======================================

[ctein]

Dear Craig,

That book **IS** somewhere in the queue, but I couldn't tell you exactly where. It's been on my mind for a couple of years; so far it hasn't gelled. That is, I have a ton of disparate facts and tips, but I don't have the coherent skeleton to hang them on. It'll come eventually.

In the meantime, I've been doing an ongoing semi-weekly series of REALLY short columns for Focal Press for their website on the subject of “The Fine Art of Digital Printing.” I figure it's a way to get the authorial juices flowing and get lots of little bits of book content written out in highly abbreviated form. if I manage to keep it up for most of the year, I'll have most of the book, even if it isn't organized into one.

It is definitely my intention to self-publish that book as a PDF. That'll let me keep the price way down; I figure I can price it at $9.95 and make it worth my while. By comparison, the paper version of POST EXPOSURE ballooned in price to over $50 at the end of its run. Paper is just so damned expensive.

I plan to hire my friend Geri Sullivan to come up with a set of design templates for me. She says she can come up with a set that will work well for both on-screen reading and for printing the book out if people want to do that. I believe her; she's really good!

That's about the only detail I've settled on, though. Will it be an expansion of POST EXPOSURE? Will it be a brand-new standalone book? Will I pay for it out of my pocket? Will I try something like Kickstarter? All unresolved.

Bridges to be crossed when the river comes into sight, which it has not yet.

pax \ Ctein
[ Please excuse any word-salad. MacSpeech in training! ]
======================================
-- Ctein's Online Gallery http://ctein.com 
-- Digital Restorations http://photo-repair.com 
======================================

[ctein]

Dear Eugene,

Oh my, my; so despite both Mike's and my admonitions, you decided to try to pass aesthetic judgment on the screen images?

That trick really doesn't work. You can't second-guess my artistic decisions or how the prints will look by looking at the illustrations here. It's not possible. Stop trying. Just look at how I changed things between the “before” and the “after” illustrations and assume, on general principle, that I did it in a proper fashion and in the proper amount.

As for any halo-ing or smudging that you think you're seeing, they do not exist in the original files and I don't even see them in the screen images here. Either there is something wrong with your display set up or you are confusing real subject detail with what you imagine artifacts from my processing to be. But the artifacts you think you see are not there. Not in the least.

pax \ Ctein
[ Please excuse any word-salad. MacSpeech in training! ]
======================================
-- Ctein's Online Gallery http://ctein.com 
-- Digital Restorations http://photo-repair.com 
======================================

[David H.]

Eugene,

These small, low resolution jpegs aren't representative of what the print will look like as Ctein noted at the bottom of his post. Although I ain't near the printer Ctein is, files of mine that look nice in print, very rarely look as good when posted on the web.

[Don Bryant]

I'm looking forward to receiving my sample print. Can't wait!

[Richard Newman]

Determining how a print should look is a subjective aesthetic issue. Two people standing side by side shooting the same subject may have very different image concepts in mind. And when the printer isnt the shooter, more judgment comes into play. I think it would be fascinating if a "comparison" of prints by Ctein and several other different expert printers, using the identical image files, could be done. The images to be printed should include a variety of types and subjects. The purpose is not to see who is "best", but to look at how different capable printers interpret the images.I suspect there would be considerable variation with some images, and little with others. What do you think? Is it feasible, and is there enough interest to try to set it up?

[Jeff]

For me the hardest part of printing is knowing when and where to make changes, and to what degree. Aesthetic judgments, that is. The 'how to' is the easy part; or at least it can be learned more easily than the interpretive part in my opinion.

It's like the surgeon explaining his fee: the scalpel cut is worth ten bucks; knowing when and where to cut is priceless. Yes, these judgments are personal and different people have vastly different tastes, but some have a much better eye than others.

I echo Ken's thoughts. It often helps to live with a print for a while, and small tweaks can often make a world of difference. I agree, too, that viewing DeCarava's prints is a special joy. In his case, I might make an exception to my comment above about the ease of 'how to.'

[Richard Tugwell]

I have to admit, I see the same effects that Eugene talks about (and on all my monitors). For example in the seascape there is a definite halo around the far headland and rocks, and on the right, where it looks as though the highlights in the water have been dodged, I see an obvious sort of shadow area. Given that small web images are worse than useless, I don't know why you don't make these available full-size for people interested. There are options to host the images externally if bandwidth is an issue. I'm sure no-one will try and steal them

[David Dyer-Bennet]

Absolutely, chasing knowledge is much more productive of good pictures than chasing gear.

I'm not as sure about formal seminars; I have no experience with them in photography. And I know people learn very differently, so the answer is unlikely to be the same for everybody. I think that many of the significant "breakthroughs" (not in the same place for each person; just getting past the places you in particular get stuck) are easier with a mentor or at least somebody "at your level" to bounce ideas around with.

This is also why looking at good prints of other people's photos is so beneficial, if you approach it in a receptive mood. I really need to do more of that.

[Daniel S.]

Thank you for giving your time to answer, and my apologies for not making my question clear enough. I meant it in the sense that if you use your memory of what you saw to dictate how the print would look, whether it gets harder to make a good print after the years have passed and the memory isn't as clear as it once was.

Of course, from your reply it seems it hasn't affected you much, if at all, and as I realized after I posted my comment, the photograph itself acts as a reminder of the scene in any case.

[hugh crawford]

Ctein,

I wonder if you could address the optimal sharpening and tone mapping as a function of print or viewing size.

I find that the best treatment of an image for a 10x15 inch print looks awful in a 50x75 inch print. A treatment that looks crisp in a small print looks lumpy in a large print and what looks crisp in a large print looks flat and undetailed in the small print

[Timprov]

Hopefully next week's column will go into more detail on your methodology for dodging and burning. I play around with this from time to time but haven't found a route to doing it competently.

[Yger]

It's always personal. To me, the "before" pictures look better.

[ctein]

Dear Richard,

Those are illusions. They're not artifacts in the image file, just in what you're seeing. Partly blog software, partly optical illusion.

You can see that for yourself if you click through to the 800px wide versions and compare them side-by-side with the in-column versions. The "halos" you think you see are much less apparent in the larger image and they are narrower. The eye creates such halos wherever there's a sharp boundary between light and dark. (If you put your face right up to the screen and look a the 800 px image, the halos mostly disappear, confirming that it's substantially a trick of the eye).

Ditto the shadow area you think you see. It looks nowhere as distinct in the 800px version; it's a trick of the compression and downsampling.

On occasions, I have hosted high res versions of illustration on my website, when I think they serve a purpose. In this case, it would not. At no resolution will these look *right*, because they're files designed to look right when printed out. We danced that waltz together back in the print sale column, remember? Screen images don't look like prints.

On that point, you need to take this line to heart: "Just as in the darkroom, it requires some real understanding of the limitations and quirks of the media." This column is about the tools and techniques needed to make good PRINTS, not screen images. These are always going to look somewhat wrong on a screen because that's not what they're optimized for.

All of which is why both Mike and I said that you should be evaluating these as comparative illustrations of techniques, not as absolute aesthetic objects.

Please, people, no more conversations about the adequacy or lack thereof of screen illos, OK? It's the communication tool we've got, warts and all, Just live with it.

pax / Ctein
==========================================
-- Ctein's Online Gallery http://ctein.com
-- Digital Restorations http://photo-repair.com
==========================================

[ctein]

Dear Stan,

I do not agree. There's nothing in new generations of software or printers that requires you to relearn your techniques. Your old methods will work just fine. What new generations of hardware and software may do is give you new tools you never had before, or easier ways of doing what you did before, but nothing obligates you to learn them.

My techniques have changed massively since the days of Photoshop 6 (yeah, 6, not CS6), but that's been my choice-- there's nothing I was doing then I couldn't do now, and little or nothing I recommend in this column that I couldn't do then. I've gotten more convenient tools, so I have learned to use them (e.g., adjustment layers are sooooooo much better for doing dodging and burning in than my old tricks and content-aware healing brushes truly rock). But I didn't have to. Neither do you.

Like Voja, I've spent a lifetime honing my printing skills. None of it ever goes to waste.

~~~~~~~~

Dear Daniel,

Oh, sorry. Yeah, that's a really, really good question. The objective answer is, "How would I know?" Of course I think I remember a scene quite well, as well as my reason for making the photo. But I only remember what I remember. Who knows if I'm right?!

Some of it's informed by objective knowledge. For instance, I know what sort of characteristic curve shape will make a night scene print like what the human eye sees. But most of it? It's really me trying to make a print that looks like what I remember seeing in the scene, and I'd be amazed if that doesn't mutate over years.

pax / Ctein
==========================================
-- Ctein's Online Gallery http://ctein.com
-- Digital Restorations http://photo-repair.com
==========================================

[AaronL]

Ctein, the shuffling you hear is TOP readers queuing up to download your pdf book. Move it to the top of your list please!!!

[richardplondon]

I expect that adding "for the Web" processing into the comparison, as distinct from the before-and after print workflow that Ctein has already demonstrated, would be both interesting and educational. Especially since we are so used to judging an image based on what we see on a computer monitor. In other words, what is different about preparing for print?

Or: if these two classes of image are considered to be substantially the same, tending to diverge only at the point of output (which is the Qimage or Lightroom "way", to take two examples) then that would be interesting to hear about too. Does whatever is beneficial for the print, generally prove beneficial for the screen display also?

[Steve Smith]

Otherwise you're trying to use a screwdriver to drive nails.

As we all know, it only works the other way round!

[Jarle Vikshåland]

Ctein, thanks for sharing your insight.

I bought the print to see what is possible wtih the M43 and I am eagerly waitning for it. I got inspired to try out the m43 cameras (in part by Kirk Turk) and bought a used Lumix G3 with the 20 mm (as an alternative to my 5D Mkii) and headed out to a nearby brook to test out the camera on some photographs that I have wanted to make or several years.

I shot from about eight in the evening, working more or less only with the afterglow of the sun (here in Oslo we are as far north as Alaska so it stays fairly light also at night now) using apertures of 8-11 and exposure from 10-30 seconds.

I worked on the print as I usually do and got a result that I was pleased with, though not quite as sharp as I had hoped.

I then read the article on how Ctein made his print. I downloaded the Infocus plugin and returned to the image. I imported it to photoshop and more or less followed your steps (plus I took some advice from Real World Image Sharpening...").
i.) Slight noise reduction with InFous, ii.) deblur and sharpening with Infocus, ii) a minor touch of sharpening using a high pass overlay .
Otherwise the same procedure with dodging, burning, curves and colour adjustment in both print.

The difference is nothing but stunning, and as pointed out you really have to see the prints side by side to see how big the difference really is. Also the A2 size print is far better than what I see on the screen.

So folks - instead of discussing, download Photoshop and the plugins as trial versions and experiment, to learn you have be willing to go through a few boxes of paper and ink cartrigdes.

PS: I am used to working in black and white in the darkroom and I am used to spending hours getting to the first really good version of a print. Digital in many ways only makes it a lot quicker to produce the 5 or 50 copies, and it has given me colour.

[Mark Roberts]

Though there's been much debate on the subject on Internet discussion boards (no, really?) the closest I've found to a comment from Adobe is that Smart Sharpen does not do deconvolution (http://forums.adobe.com/thread/748923?start=0&tstart=0), though this is from Nov. 2010.

I've had great results with Focus Magic and Focus Fixer in the past but neither seems inclined to produce a 64-bit version (or indeed any kind of update).

[mike plews]

Even the very best photo books don't tell you all that a good reference print will.
All the techinical information in the world won't get you to the top of the mountain unless you educate your eye as to what a good print looks like by experincing one directly.
If you can afford a nice print consider it tuition. If not, hit the museum on a regular basis (do this anyway).
Back in the mid 1960's I got really lucky. While in high school I worked for a local studio as a printer. The owner had several Adams prints and even an Edward Weston. Remember this was the 60's and vintage Adams and Weston prints weren't selling for the price of a couple of shares of Berkshire A stock.
The owner was good about teaching me what he knew which was considerable but walking out of the dark with a print in the tray and looking at a print of the Merced river was a humbling experience.
That thing mocked me. It set a standard I still can't meet.

[Taran]

Just a word about home vs outsourcing...

Prints you make at home should be better than any professional prints provided you have a professional printing device (given time and effort).

This is because professional printers are at a significant disadvantage to a hobbyist who is custom tailoring his prints to his needs. The professional has to have a baseline output of SRGB 8 bit color because that is how 99.9% of his prints are produced. When you need a 16 bit Pro Photo TIFF output, there is no way the professional can compete with someone like me, who does 16bit Pro Photo to boutique papers, all day, everyday.

Professional printing, in general, is a misnomer, because the technologies have changed so rapidly in the last 5 years, only a committed hobbyist could keep up. A professional has to make money, but a hobbyist will hemorrhage cash to get his ideal result via experimentation.

I see many prints in museums and private collections that are fine art pigment, none of them are as good as the prints I can produce, because they were made by "professionals".

[Bob Rosinsky]

No doubt Ctein is a master darkroom technician. However, his discussions on digital printing are rudimentary. For example, instead of burning and dodging, it is far more precise as well as nondestructive to create curve adjustment layers and then selectively "paint" the corrections onto a mask. Understanding color space and intelligent use of channel corrections is another area absent from his discussions.One of the greatest aspects of Photoshop is the ability to use LAB for powerful and rapid color correction. LAB is also useful for sharpening luminance detail without corrupting color data. Moving from LAB back to sRGB or aRGB is nondestructive. With all due respect, I sometimes feel tht Ctein's Photoshop tutorials hinder rather than help.

[Edward Taylor]

Ctein -

I enjoyed reading about being a good printer, but in a related note - I received the print you did of the bridge and moon today. All I can say is - WOW! Talk about actions speaking louder than words ...

Ed

[MM]

Thanks much to Ken for his comment -- especially vis-a-vis what is a good print vs. what the Internet experts say a good photo is -- and even more for his DeCarava example. To my eye DeCarava is as good as it gets when it comes to b&w tonality, and that photo is as good as DeCarava gets (one doesn't usually use the word "gutsy" when discussing darkroom work, but that word accurately describes Roy's printing choices; I think his print-inspection light had to be a candle or a 5-watt bulb). He discussed that photo in Roy DeCarava Photographs, in a quote that is reprinted on the Museum of Contemporary Photography website:

"This photograph was taken at a dance of a social club at the 110th St. Manor at Fifth Avenue. It is about the intermission where they had entertainment and the entertainment was two dancers who danced to jazz music. That's what this image is all about; it's about these two dancers who represent a terrible torment for me in that I feel a great ambiguity about the image because of them. It's because they are in some ways distorted characters. What they actually are is two black male dancers who dance in the manner of an older generation of black vaudeville performers.

"The problem comes because their figures remind me so much of the real life experience of blacks in their need to put themselves in an awkward position before the man, for the man; to demean themselves in order to survive, to get along. In a way, these figures seem to epitomize that reality. And yet there is something in the figures not about that; something in the figures that is very creative, that is very real and very black in the finest sense of the word. So there is this duality, this ambiguity, in the photograph that I find very hard to live with.

"I always have to make a decision in a case like this -- is it good or is it bad? I have to say that even though it jars some of my sensibilities and reminds me of things that I would rather not be reminded of, it is still a good picture. In fact, it is good just because of those things and in spite of those things. The picture works."

[Mark K Lough]

With regard to the references to "deconvolution sharpening". It set me off to do some Googling. After reading several of the posts, it seems that deconvolution is the attempt(s) to de-construct the effects of the filter placed over bayer array image sensors. That would be all digital cameras other than Sigma I believe.
Has there been a comparison made of images produced with digital cameras versus images made on film and scanned, with regard to the need to sharpen?
Maybe it's too simple, but it seems to me that eliminating the filter-over-the -sensor convolution of a recorded image would be just what happens when one shoots film.
Anyone done that comparison?

[John Rodriguez]

It is not currently possible to get the same level of control in Lightroom, even LR 4 is way off. It's fine for organizing images and doing some intitial "development", but for making prints that match your intent it just doesn't have the tools. However, note that I just said "match your intent". The biggest obstacle to making a great pring is knowing what you want. Once you have that it makes learning image editing software (any kind) much easier.

A few of the things I consider to be bread and butter tools that LR is missing -

- Separate control of all 10 channels (RGBLABCMYK). LR doesn't even provide access to CMYK or LAB globally.
- Layers, and in turn layer blending modes and the uber useful Blend-If function
- Channels
- Apply Image. The best color images have an excellent black and white image behind them. There are several tools that will help you get there, Apply Image is the tool I use most to create BW luminosity layers for my color images.
- Smart Objects.
- Select Color command.

It's important to note, that all the image editing stuff in LR is available in PS through ACR. One thing I like to use ACR for is deconvolution sharpening, the tools are a bit better then what's available in Smart Sharpen (CS5 at least, haven't tried 6). Also, the new Highlight and Shadow tools in LR4 look very powerful. One caveat though - LR/ACR doesn't play all that nice with uber large files unless you have a TON of RAM (I regularly work on 4x5 scans that start at ~600MB).

[Richard]

Ctein,

I need to do some reading on "deconvolution sharpening". It's new to me. I have used a high pass filter in PS (either globally or locally in a PS mask). Have you done that or does the deconvolution sharpening give a better result with fewer artifacts, which, at least in my limited experience with outdoor images such as aspen leaves gives nice edges without the problems associated with "too much sharpening".

About the book, please, please do an eBook. It offers so much better opportunities to to link to additional material and to update it in the future without the hassle of printing a new physical edition. (That said, physical books are wonderful. I treasure "The making of 40 Photographs."

I might suggest looking into the ePub format rather than a PDF. I am told that it can be view in either portrait or landscape modes and has other capabilities not present in a PDF document format. Either way, please do it. I will await its availability with great anticipation.

Cheers,
Richard

[hugh crawford]

Mark,

"Maybe it's too simple, but it seems to me that eliminating the filter-over-the -sensor convolution of a recorded image would be just what happens when one shoots film."

Well but then you have the scanner convolutions to deal with reminds me of a certain lens company whose camera lenses and enlarging lenses that both had the same barrel distortion and vignetting thus canceling each other out.

[Jeff Markus]

Being able to directly view prints by an excellent print-maker is an invaluable tool to improve your own print-making. Back in the day, I too made my living as a custom printer. I was extremely fortunate in that, early on in my first printing job, the local art museum opened up a good-sized gallery devoted solely to photography. The initial show was an Edward Weston career survey, and the second was Ansel Adams' complete Museum Set. They also had a set of Moonrise prints showing the evolving interpretation of that negative. I had the rare privilege of being able to spend many afternoons studying those prints, and then applying those lessons that very night in the darkroom.

It was this experience that made me realize the huge importance of local contrast. Adams had begun his career making very soft-toned prints, then gradually hardened up on his tonality. A good part of that was his use of a gradually increasing local contrast. The Museum Set was done at the end of this process, and the prints really "popped". They had a three-dimensionality to them; they were objects themselves, not merely a flattened two-dimensioned depiction of an object. I believe that it's this sense of dimensionality that makes a print come alive, versus one that sits there flat and barely breathing.

[ctein]

Dear Mark R. (and others),

Oh, that's interesting. Well, I would never argue with Chris. He's the expert. Apparently I was informed wrong. Smart Sharpen is not a deconvolution routine.

I'm quite surprised; the results it produces to look a lot like deconvolution. But it's hard to argue with facts.

It still works a lot better than ordinary dumb sharpening. And at the small radii that I typically do sharpening (0.3-2 px), I have to say it requires pixel peeping to see the superiority of one method over another. On the other hand, when one has determined that one method is superior, no reason not to stick with it.

In general, deconvolution or Smart Sharpen works a lot better than edge-enhancement “sharpening”, whether done with ordinary sharpening filters or with high Pass filter tricks. You get fewer artifacts and you get better sharpening of low contrast details, not just obvious edges.

~~~~

Dear Mike P.,

This was the rationale behind our 'Steamed Editor's brilliant idea to do print sales at rock-bottom prices (which he started when he was editor of PHOTO Techniques magazine). The fundamental intent was to get excellent prints into people's hands at prices they could afford so they would know what good printing really looked like.

That they might get some nice art as well was a welcome fringe benefit.

~~~~

Dear Taran,

Why do you say that a professional printer has to operate in sRGB 8-bit color? I don't. I don't know any other professional printers who do. Not saying they don't exist, but it's hardly de rigueur or even the norm.

There are commercial photofinishing services that will only accept files in that form for printing, but they aren't doing custom/professional printing as were talking about it in this column. All they're doing is providing the output service. To make best use of them, one still have to do all the work onerself, that I outlined in this article; you're just not generating the final hard copy.

So, yes, if one is letting others do the physical printing, they may be at a disadvantage… but they are still having to do the hard lifting.

By the way, being limited to sRGB color space does make a difference in many (not all) cases; Adobe RGB is a much better match to printers (and I work in Pro Photo, like you). But, so far as I can tell, once you've massaged your photograph into shape, it makes no difference in the quality of output whether it is an 8-bit file or a 16-bit file. Both Dave Polaschek and I are still trying to find a real-world photograph where 16 bit printing workflow produces any visible difference. I mean, you'd expect it to, but we can come up with a case where we can see it.

Life continues to be a mystery [grin].

pax \ Ctein
[ Please excuse any word-salad. MacSpeech in training! ]
======================================
-- Ctein's Online Gallery http://ctein.com 
-- Digital Restorations http://photo-repair.com 
======================================

[ctein]

Dear Bob,

My, but you certainly make an awful lot of assumptions from a summary article. An overview article, by definition, doesn't get into specific technique; there isn't room in the word count, and it distracts from the overall message.

If you reread more carefully, you will see that I did not say how I did dodging and burning in. In fact, I was trying to be as non-program-specific as possible.

Also, in fact, I do dodging and burning in exactly as you described, and as I've written about on numerous occasions. Apparently you missed those. But here are two examples from my website, as evidence:

http://photo-repair.com/Chapter_10-Examples_(from_Digital_Restoration_1st-ed).pdf
http://photo-repair.com/DRBookPromo/DR_Excerpt3.htm

Regarding various color spaces, I have also written several articles on the subject, and Lab space is neither the devil incarnate nor the panacea that folks at the extrema (vis Martin Evening and Dan Margulis ) seem to think. There are, in fact, useful tricks specific to RGB or Lab space that are extremely difficult if not impossible to duplicate in the other space.

Oh, and by the way, since we seem to be into picking nits, conversion between the spaces **IS** destructive. it should not be done for no good reason. But when there is a reason, the improvements to the photograph far outweigh the minor losses of information.

In the future, you might want to not make assumptions about how much I know based on what you DON'T see in an article.

With all due respect. Of course.

pax \ Ctein
[ Please excuse any word-salad. MacSpeech in training! ]
======================================
-- Ctein's Online Gallery http://ctein.com 
-- Digital Restorations http://photo-repair.com 
======================================

[ctein]

Dear Mark L.,

There is a practical reason for using deconvolution sharpening less with film scans-- on average, film scans are noisier than digital photographs, due to the film grain. Since deconvolution sharpening improves all fine detail and it doesn't know the difference between noise/grain and subject detail, it makes it less desirable and more problematical to use well on noisier images. It's almost inevitably part of my basic tool kit for dealing with digital photographs, but I only use it sometimes with film. In fact, I'm more likely to use it when the film photograph is really, really sharp to begin with. In that case, a proper mix of sharpening and noise reduction does produce a result that is both sharper and finer grained than the original.

I wrote an article about this for PHOTO Techniques years and years ago. If you were a subscriber, you should be able to dig it up.

pax \ Ctein
[ Please excuse any word-salad. MacSpeech in training! ]
======================================
-- Ctein's Online Gallery http://ctein.com 
-- Digital Restorations http://photo-repair.com 
======================================

[Mark K Lough]

Ctein

Thanks for the improved resolution. I look forward to reading (for likely several months) your latest book Digital Restoration From Start To Finish, Second Edition, which is on its way to me now.
I've done copy and some restoration photography film and darkroom style for quite a while, and am really ready to partake of your experience via the new book.

[ctein]

Dear Timprov,

I wrote a short piece about how to do dodging and burning in for the Focal Press website:

http://masteringphoto.com/the-fine-art-of-digital-printing-layers-are-the-way-to-dodge-and-burn/

I've been doing a series of shorts for them on the fine art of digital printing (or the art of fine digital printing, as you prefer). There should be 16 of them online, but finding them all is difficult; it's kind of a disorganized site. These two URLs turn up most of them:

http://masteringphoto.com/author/ctein/
http://masteringphoto.com/?s=ctein&x=11&y=15

See you in a bit over a month!

pax \ Ctein
[ Please excuse any word-salad. MacSpeech in training! ]
======================================
-- Ctein's Online Gallery http://ctein.com 
-- Digital Restorations http://photo-repair.com 
======================================

[ctein]

Dear Hugh,

There are two different questions in there. I think I'll tackle them in two comments for clarity's sake.

Regarding tonality and contrast, it's long been known to fine darkroom printers that smaller prints look more contrasty than larger ones. A negative that prints well on Grade 2-1/2 paper in 8 x 10 or 11 x 14 size will more likely look good on Grade 2 paper in a 4 x 5 print and Grade 3 paper in a 20 x 24 print. Part of it is likely nonlinear behavior of the printing materials, and part of it is likely human vision characteristics: we more easily pick up on tonal differences over small angular areas than large ones, and we tend to view larger prints from a position that makes them cover a greater angular area in our field of vision.

The effect seems much less strong in digital printing; something that looks good to me at 11 x 14 size still looks good printed out at 40" x 50". But really small prints look snappier and it's very much true for web images. Makes it additionally hard to produce a screen image that looks anything like print.

You can get a bit of a sense of this for yourself if you go to my website and look at the thumbnails and then click through them to the “full screen” images. Both are generated from exactly the same file in the same way, just scaled to fit, but the thumbnails look contrastier (since they're just thumbnails, I don't care enough to fix that).

pax \ Ctein
[ Please excuse any word-salad. MacSpeech in training! ]
======================================
-- Ctein's Online Gallery http://ctein.com 
-- Digital Restorations http://photo-repair.com 
======================================

[ctein]

Dear Hugh,

On to the second part of your question…

As for optimal sharpening, my experiences and results are substantially at odds with that of many other printers. The majority of the time I don't need to “sharpen for output” and when I do I very rarely need to use different kinds and levels of sharpening for different size prints. I'm not at all sure why my experience is so different from what other people report, but I have a most tenuous theory (barely half-baked):

I think most sharpening for output is a hack, it's a good-enough trick for fooling the printer into producing better output and its quick and easy, but it's not really the right answer. The evidence for that is when you pixel-peep the file at 100% on your monitor, It seems to be mostly about greatly exaggerating edges and those output-sharpening hacks look really lousy! They have all sorts of warts that aren't visible in the output, but clearly they influence how stuff gets printed. So it's not too surprising that if you print much larger you see those warts in all their glory, and if you print much smaller nonlinearities in the rendering engines may not make things scale the way you expected.

I recently had a little experience with this that sorta supports my hypothesis. I've been doing all the printing (both for paper and for the web) for the Bernard Lee Schwartz Foundation for 30 years; they are my biggest and longest-duration client: http://www.bernschwartz.org . About a half dozen years ago, it was strictly dye transfer printing for them; then we started to move into digital and now it's entirely digital. My standard digital print size for them has a 16" x 20" image area or 20x24 paper (not much different from the dye transfers). Recently I printed a retrospective show for the Jerusalem Museum for them. The museum wanted a few large prints for accent, so five of the photographs also got printed as up to 44 x 55 prints.

Two of those photographs I could print with very little change from the same files I used for the 20 x 24 prints. Oh there was a small amount of corrective work; very minor stuff that wasn't visible in a small print but was annoying in a large one. No changes in sharpening or things like that. One of them required a moderate amount of reworking and the other two I had to go back to square one and start all over. The difference was that on the two that printed with no trouble, they were good down to the single pixel level; I gotten the noise reduction, the sharpening, the local contrast enhancement all right from the get go. In the two really difficult ones, I'd made some serious compromises along the way (for technical and artistic reasons I shan't go into) that amounted to hacks. If you took them up to 100% on the screen, you could see they looked lousy on the pixel level; they just printed really nicely as 20 x 24's . Taking them up to 250% of that size exposed all my crap work. So I had to go back and do them **right** .

Having done them right, big, they print very nicely small, not so by the way. My original hackwork didn't make a better-looking print than doing it up properly, it was just good enough.

Anyway, that's been my experience to date, for what very little it's worth, and I'm not sure it's worth a lot.

pax \ Ctein
[ Please excuse any word-salad. MacSpeech in training! ]
======================================
-- Ctein's Online Gallery http://ctein.com 
-- Digital Restorations http://photo-repair.com 
======================================

[Timprov]

Dear Ctein -

Having read it but not tried it, I expect that is exactly what I needed, thank you.

Looking forward to getting to show you my prints rather than just the other way around.

[Kenneth Tanaka]

@ Ctein: I did not know about your work for the late Bern Schwartz. He did some lovely portrait work in just four years. I'm sure those prints look fabulous. I expect to be in London sometime later this year. Perhaps I can schedule a viewing of some of these works.

[Ctein]

Dear Ken,

Hmmmm, did I never write about that? I'll have to check when I get home. If not, it'll be next week's column, prolly.

Anyways, yes, everything in the British Portait Gallery's collection is stuff I printed, dyes and digitals, and scanned for them. The BLS Foundation also donated all the original negatives to their collection.

pax / Ctein

[John Gillooly]

Ctein, do you have a website or contact information regarding using you as printer. i have long been interested in having a professional printer work on and print some files. i can't think of a better way to SEE what a real print should look like than comparing what i would do versus what an actual printer would do. I would be interested in this service and getting fee structure, etc.

John
[email protected]

[ctein]

Dear John,(and anyone else who might care),

http://ctein.com/customprinting.htm

pax / Ctein